karanjkar99 / Tsa

Time Series Analysis

TIME SERIES ANALYSIS

Time Series Analysis is a series that often presents itself in day to day situations and requires forecasting, be it stock prices, weather, etc.

In this project, we will forecast the "Closing Price " of Yes Bank (BSE) stock using various techniques.

The first step is to acquire the data that we perform using the Yahoo finance library, and then we present the Dickey-Fuller test, which is used to measure the stationarity of any Time Series.

After this step, we plot various graphs to ease the analysis using visual representation. Which is followed by multiple plots drawn to represent -

• Seasonality
• Trend
• Correlation

These plots are followed by sampling the data into two parts the first 80% for training the model and the remaining for testing the model. We perform curve fitting using the ARIMA model based on minimisation of SMAPE(squared mean absolute percentage error).

To conclude, the predictions are adequately accurate, but as the model is dynamic, its value will vary in different time frames.